OBJECTIVE To examine the mechanisms by which weight loss improves glycemic control in overweight subjects with NIDDM, particularly the relationships between energy restriction, improvement in insulin sensitivity, andregional and overall adipose tissue loss. RESEARCH DESIGN AND METHODS Hyperinsulinemic glucose clamps wereperformed in 20 subjects (BMI = 32.0 ± 0.5 [SEM] kg/m2, age = 48.4 ± 2.7 years) with normal glucose tolerance (NGT) (n = 10) or mild NIDDM (n = 10) before and on the 4th (d4) and 28th (d28) days of a reduced-energy (1,100 ± 250 [SD] kcal/day) formula diet. Body composition changes were assessed by dual energy x-ray absorptiometry and insulin secretory changes were measured by insulin response to intravenous glucose before and after weight loss. RESULTS In both groups, energy restriction (d4) reduced fasting plasma glucose (FPG) (ΔFPG: NGT = −0.4 ± 0.2 mmol/1 and NIDDM = −1.1 ± 0.03 mmol/1, P = 0.002), which was independently related to reduced carbohydrate intake (partial r = 0.64, P = 0.003). There was a marked d4 increase in percent of insulin suppression of hepatic glucose output(HGO) in both groups (ΔHGO suppression: NGT = 28 ± 15% and NIDDM = 32 ± 8%, P = 0.002). By d28, with 6.3 ± 0.4 kg weight loss, FPG was further reduced (d4 vs. d28) in NIDDM only (P = 0.05), and insulin sensitivity increased in both groups (P = 0.02). Only loss of abdominal fat related to improvements in FPG (r = 0.51, P = 0.03) and insulin sensitivity afterweight loss (r = 0.48, P = 0.05). In contrast to insulin action, there were only small changes in insulin secretion. CONCLUSIONS Both energy restriction and weight loss have beneficial effects on insulin action and glycemic control in obesity and mild NIDDM. The effect of energy restriction is related to changes in individual macronutrients, whereas weight loss effects relate to changes in abdominal fat.